Risk-Sensitive Submodular Optimization
نویسنده
چکیده
The conditional value at risk (CVaR) is a popular risk measure which enables risk-averse decision making under uncertainty. We consider maximizing the CVaR of a continuous submodular function, an extension of submodular set functions to a continuous domain. One example application is allocating a continuous amount of energy to each sensor in a network, with the goal of detecting intrusion or contamination. Previous work allows maximization of the CVaR of a linear or concave function. Continuous submodularity represents a natural set of nonconcave functions with diminishing returns, to which existing techniques do not apply. We give a (1− 1/e)-approximation algorithm for maximizing the CVaR of a monotone continuous submodular function. This also yields an algorithm for submodular set functions which produces a distribution over feasible sets with guaranteed CVaR. Experimental results in two sensor placement domains confirm that our algorithm substantially outperforms competitive baselines.
منابع مشابه
Some Results about the Contractions and the Pendant Pairs of a Submodular System
Submodularity is an important property of set functions with deep theoretical results and various applications. Submodular systems appear in many applicable area, for example machine learning, economics, computer vision, social science, game theory and combinatorial optimization. Nowadays submodular functions optimization has been attracted by many researchers. Pendant pairs of a symmetric...
متن کاملSupplemental Material: Risk-sensitive submodular optimization
Proof. We can establish the result for fixed x using the proof of Ohsaka and Yoshida. We have via taking c = L in their Lemma 4.4 that for any fixed x, |CVaRα(x) − ̂ CVaRα(x)|≤ with probability at least 1 − δ by taking s = Θ ( M2 2 log 1δ ) samples. Note that we cannot directly take union bound because the set of x ∈ P is not finite. Instead, we take a uniform grid of ( L1d )n points containing ...
متن کاملForthcoming in Mathematical Programming MAXIMIZING A CLASS OF SUBMODULAR UTILITY FUNCTIONS
Given a finite ground set N and a value vector a ∈ R , we consider optimization problems involving maximization of a submodular set utility function of the form h(S) = f (∑ i∈S ai ) , S ⊆ N , where f is a strictly concave, increasing, differentiable function. This function appears frequently in combinatorial optimization problems when modeling risk aversion and decreasing marginal preferences, ...
متن کاملPolymatroids and mean-risk minimization in discrete optimization
We study discrete optimization problems with a submodular mean-risk minimization objective. For 0-1 problems a linear characterization of the convex lower envelope is given. For mixed 0-1 problems we derive an exponential class of conic quadratic inequalities. We report computational experiments on risk-averse capital budgeting problems with uncertain returns.
متن کاملStochastic Submodular Maximization: The Case of Coverage Functions
Stochastic optimization of continuous objectives is at the heart of modern machine learning. However, many important problems are of discrete nature and often involve submodular objectives. We seek to unleash the power of stochastic continuous optimization, namely stochastic gradient descent and its variants, to such discrete problems. We first introduce the problem of stochastic submodular opt...
متن کامل